The%20Equatorial%20Cold%20Tongue%20Bias%20in%20a%20Coupled%20Climate%20Model

1
The Equatorial Cold Tongue Bias in a Coupled
Climate Model

• Vasu Misra1, Larry Marx1, M. Brunke2 and X. Xeng2
• 1 Center for Ocean-Land-Atmosphere Studies
  (COLA),
• 2 Department of Atmospheric Science, University
  of Arizona

2
Variance of MJO mode (30-70days wns 1-6)
Courtesy Jia-Lin
3
Why is one-day coupling interval so popular?
4
Why is one-day coupling interval so popular?

• Most convenient
• Best compromise of available resources
• Practical limitation and reconciliation to
  approximations
• Noise is important

5
Conducted experiments

• CON Integration of control COLA coupled model
• E2D Coupling frequency changed to 2 days
• E3D Coupling frequency changed to 3 days
• E3H Coupling frequency changed to 3 hours
• ESKIN AGCM interacts with skin SST at every time
  step
• All experiments are run for 50 year period from
  well spun-up ocean and land surface initial
  conditions.

6
ESKIN

• The coupling interval between OGCM and AGCM is
  once a day. Therefore, the bulk SST (10m) is
  updated once a day.
• However skin SST is updated at every time step of
  the AGCM.
• Skin SSTf(TB,Q,u,?t)

7
COLA coupled model

• Radiation (CAM)
• PBL (NCEP non-local)
• Convection (NASA RAS)
• Land surface (COLA SSiB)
• Resolution T62L28
• Ocean (MOM3)

8
Climatological monthly mean errors of SST over
equatorial Pacific
9
SST
Climatological Dec-Jan-Feb differences. Only
significant values are shaded
10
Velocity potential _at_ 200 hPa
Climatological Dec-Jan-Feb differences. Only
significant values are shaded
11
Velocity potential _at_ 200 hPa
Climatological Dec-Jan-Feb differences. Only
significant values are shaded
12
1000hPa winds and surface pressure (Pa)
Climatological Dec-Jan-Feb differences. Only
significant values are shaded
13
The first EOF of SST
14
(No Transcript)
15
Regression of zonal wind stress anomalies on
Nino3 SSTA reconstruced from EOF1
16
Computational time time

• On the same machine, for the same number of
  processors
• E3H38 more than CON
• ESKINCON
• E2D3 less than CON
• E3D6 less than CON

17
Conclusions

• Diurnal coupling interval even with coarse AGCMs
  and OGCMs has an impact.
• Skin SST interactions seem to damp the convection
  over the oceans-may have huge implications in
  models that have strong ENSO.
• Approach to tropical bias has to come from
  incremental improvements.

18
Zonal wind stress
Climatological Dec-Jan-Feb differences. Only
significant values are shaded